1. Field of the Invention
The present invention relates generally to the fiberization of glass or other thermoplastic materials and relates more particularly to fiberization techniques wherein the molten material to be fiberized is centrifugally converted by a rapidly rotating spinner into a multiplicity of glass streams which are attenuated into fibers by a concentric annular gaseous blast from an internal combustion burner adjacent the periphery of the spinner directed perpendicularly to the centrifugal stream, such a fiberization technique being herein referred to as "centrifugal blast attenuation". The fibers, after being sprayed with a binder, are collected on a foraminous conveyor in the form of a blanket or mat, which is then passed through a curing oven.
2. Description of Prior Art
The centrifugal blast attenuation glass fiberization technique generally described above has been used industrially for many years in the production of glass fiber insulation products, and a substantial percentage of glass fiber insulation manufactured at the present time is produced utilizing this technique. Details of various forms of this process are disclosed for example in U.S. Pat. Nos. RE 24,708 2,984,864, 2,991,507, 3,007,196, 3,017,663, 3,020,586, 3,084,381, 3,084,525, 3,254,977, 3,304,164, 3,819,345 and 4,203,745.
In carrying out this technique, substantial amounts of heat energy are required, first for heating the glass into a molten state, and secondly for producing the attenuating blast. The uncertain availability and high cost of energy have created an increasing demand for glass fiber insulation products, while the same factors have caused a substantial increase in the cost of producing such products.
Efforts have accordingly been made to improve the efficiency of the described fiberization process or to utilize alternate fiberization techniques. For example, some glass fiber production has in recent years been carried out utilizing a purely centrifugal fiber attenuation, primarily to avoid the energy requirements of the blast attenuation technique. Such a process is disclosed for example in U.S. Pat. No. 4,058,386.
Centrifugal stream formation coupled with blast attenuation as generally described above remains a preferred technique however, both because of the excellent quality of the fiber blanket obtained therewith as well as the fact that a substantial portion of the insulation industry is equipped at present with apparatus for carrying out such a process. It accordingly follows that any improvement in this technique would be of significant industrial importance. As will be understood from the following disclosure, the present invention provides marked improvements in centrifugal blast attenuation fiberizing techniques with respect to product quality, production rate, and operating costs.
Inasmuch as glass fiberization is in practice an extremely complex technique characterized by a large number of variable parameters, many of the details of known techniques need not be included herein, reference being made to the above patents for such disclosures. However, certain limited aspects of the prior art will be considered, especially concerning those factors respecting which the present invention departs substantially from prior practice.
Among the many variables to be considered, the construction of the spinner and the speed at which it rotates are of particular importance in successfully carrying out a centrifugal fiberization process. In addition, the diameter of the spinner, the size, number and arrangement of the orifices in the peripheral spinner wall, the alloy from which the spinner is made as well as the shape of the spinner wall, the distribution of molten glass to the interior spinner wall and the control of the temperature of various portions of the spinner assembly and the glass flowing therewithin are factors which must be carefully considered.
In the centrifugal blast attenuation process, the blast temperature and velocity, as well as the placement of the blast nozzle and direction of the blast with respect to the spinner wall are important to an optimization of the fiber attenuation. Spinner life is an important factor, particularly in view of the relatively short life of this type of spinner and the extremely high cost of spinner replacement.
The spinners used in early centrifugal blast attenuation equipment were typically of a diameter of about 200 mm and the peripheral wall thereof included typically 4,000 to 6,000 holes through which the molten glass passed to form the primary glass streams subjected to attenuation by the annular blast. It was perceived at an early date that for a spinner of given size and construction, the output or pull rate, conventionally expressed in terms of the weight in tons per day of produced fiber, could be increased only at the expense of a corresponding decrease in fiber quality. It was further perceived that there were practical limits to the pull rate per spinner orifice for maintaining acceptable fiber quality, the maximum rate per orifice ranging between about 1 and 1.4 Kg/day. Nonetheless, the economic demands for increasing production of a given line usually resulted in an increase in pull rate despite the deterioration in product quality. The term "quality" in this sense refers to the product weight per unit of area for a given thermal resistance and nominal product thickness. A lower quality product would hence be a heavier product although with the same insulating value as the better quality product. The lower quality product is thus lower in quality not only since it has a higher density, but also in the sense that it is inherently a more expensive product, requiring more glass for a given area, and is thus more costly to manufacture.
In an effort to increase the output of a spinner of given diameter, the number of holes in the peripheral wall of the spinner was increased. Although some increase in pull rate was achieved, there are practical limits of orifice density controlled by factors such as the necessity of maintaining discrete glass streams emerging from the periphery of the spinner and manufacturing problems. Similar considerations limit the degree to which the spinner peripheral wall can be increased in height to increase its area.
Since the pull rate per orifice, orifice density, and height of the spinner wall could not be further increased without sacrificing fiber quality below acceptable limits, efforts to increase the pull rate were directed toward increase of the spinner diameter, initially to 300 mm and more recently to 400 mm. Although each such increase in diameter produced some increase in pull rate and/or an improvement in fiber quality, the improvements were modest in comparison with those of the present invention.
Another limiting factor is the centrifugal acceleration produced by the high rate of spinner rotation. Although substantial centrifugal forces are required to produce the necessary flow of molten glass through the spinner orifices and to thereby form the primary glass streams, high centrifugal forces foreshorten the life of the spinner.
Since spinner life is substantially inversely proportional to the spinner centrifugal acceleration forces, it has heretofore been considered desirable to restrict rotational speeds of the spinner as much as possible in an effort to extend the spinner life.
Due to the detrimental effects of higher centrifugal acceleration on spinner life and the uncertain effects of higher peripheral speeds on fiber attenuation, the conventional wisdom when increasing spinner diameter has been to decrease or refrain from increasing the centrifugal acceleration and to hold peripheral velocity within a range known to give satisfactory attenuation.
A further factor is of importance, namely the fineness (average diameter) of the fibers. It is well established that for a given density of fiber mat layer, the finer the fibers, the greater the thermal resistance of the layer. An insulating product comprising finer fibers can accordingly be thinner with the same insulating value as a thicker product of coarser fibers. Or, likewise, a product of finer fibers can be less dense than one of coarse fibers of the same thickness and have the same insulating value.
Since sales of insulation products are usually based on a guaranteed thermal resistance (R value) at a nominal thickness, the fiber fineness is an important factor determining the weight of the product per unit of area, known as the basis weight, a product of finer fibers having the lower basis weight and hence requiring less glass and enjoying manufacturing economies.
From an economic standpoint, however, fiber fineness, as with other factors, is normally considered to be a compromise since the attainment of finer fibers is thought to flow principally from higher blast velocities and from the use of softer glass compositions. Increasing the blast velocity results in a direct increase in energy costs, and softer glasses typically require ingredients which are expensive and which, further, usually have undesirable pollutant characteristics.
Fineness, which can be expressed in terms of fiber diameter, in microns, representing the arithmetic mean value of measured fiber diameters, is also conveniently expressed on the basis of a fiber fineness index, or a "micronaire" determination, the latter being a standard measuring technique in the glass wool industry wherein a predetermined mass or sample, for example 5 grams of the fibers, is positioned within a housing of a given volume so as to form a permeable barrier to air passing through the housing under a predetermined pressure and the measurement is made on the air flow through the sample. The measurement thus made depends on the fiber fineness.
In general, the finer the fibers the more resistance offered to the passage of air through the sample. In this manner an indication is given of the average fiber diameter of the sample. The fineness of typical blast attenuated centrifugal glass fiber insulation products ranges from fine types (i.e. micronaire 2.9 (5 g); average diameter 4 .mu.m) to relatively coarse types (i.e. micronaire 6.6 (5 g); average diameter 12 .mu.m).
The insulating value of a blanket of fibers is dependent to a limited although significant degree on the lay-down of the fibers on the collecting conveyor, which determines the orientation of the fibers in the insulation product. The thermal resistance of a fiber blanket will vary depending on the direction of orientation of the fibers to the measured heat flow, the resistance being greater when the fibers are oriented perpendicular to the direction of heat transfer. Accordingly, to maximize the thermal resistance of an insulating blanket, the fibers should be oriented to the maximum degree possible in an atti de parallel to the collecting conveyor and the plane of the blanket formed thereon. Because of the extreme turbulence generated above the collecting conveyor by the decelerating fibers and gaseous currents, there is very little that can in this area of the fiberizing process being directed toward achieving a relatively uniform distribution of fibers across the width of the conveyor.
In an effort to increase fiber production still further while maintaining or possibly improving the quality of the fibers, experiments were undertaken with still larger spinners having diameters of 600 mm and over. Surprising improvements in pull rate and/or quality were achieved although for reasons some of which at the present time are not entirely clear. Particularly unexpected were improvements in the fiber quality which exceeded forecasts by 10-25% depending on the pull rate utilized.
In making the transition to the large size spinner, th spinner rotational speed was reduced to provide centrifugal accelerating forces on the glass and spinner wall within conventional limits such that the glass feed through the orifices, as well as the stresses placed on the spinner wall structure, would not depart significantly from prior practice. It was feared that the larger spinner diameter would, at the rotational speed necessary to produce such centrifugal force, result in an unacceptably high peripheral speed of the spinner with a consequent degradation of fiber formation and attenuation. Surprisingly, the fiberization was not adversely affected by the higher peripheral speed, but, to the contrary, was actually improved as evidenced by the improved fiber quality and/or pull rate as compared with lower peripheral speeds. Furthermore, it has been found that the larger spinner diameter significantly improves the operating economy of the system.